A Study on Advanced Network System of Ubiquitous Small-scale Aerial and Ground Robots for High Performance Prevention of Disaster

近年来，多旋翼直升机型自主无人机被用于航空摄影和航空测量。此外，多旋翼直升机还可用于建筑物维护、安全、救援等多种用途。在不久的将来，这些技术将渗透到我们的生活中。然而，如果在没有指导方针和安全保障的情况下广泛传播，可能会发生严重事故。因此，制造商需要能够向用户展示故障模式的适用范围和行为。为了说明这一点，需要能够用于分析行为模型参数变化和硬件故障的解析模型。我们的研究小组从2007年开始研究多旋翼直升机。2010年，我们开始开发具有自己设计的实用多旋翼。我们的研究目标是在本项目中实现无处不在的小型空中和地面机器人的先进网络系统，以实现高性能的防灾。2011年，我们在自主控制方面取得了更大的进展，并于2012年推出了多旋翼机联盟。如上所述，要实现安全的多转子，需要建立解析模型。在本研究中，我们建立了一个适用于一般多转子的分析模型。通过构建控制系统，对模型进行了验证。我们构建了一个仿真系统，可用于操作员培训和控制系统验证。此外，该模拟器还可以模拟转子故障和传感器故障。实际应用表明，该仿真器可用于分析转子故障行为，并可用于培训操作人员。虽然多旋翼直升机主要用于室外，但也有望用于室内信息采集。飞行机器人在室内等无gps环境下的自主导航要求飞行机器人能够利用外部传感器估计位置。虽然激光扫描仪主要用于室内飞行研究，但由于传感器质量较大，阻碍了小尺寸机器人的发展。因此，在本研究中，我们开发了一种使用四个红外（IR）传感器实现室内自主飞行的轻型飞行机器人。然而，基于红外传感器的定位存在以下问题。首先，很难使用靠近墙壁的红外传感器，因为这样做会在使用传感器输出电压计算距离时产生错误的结果。第二个问题是空间分辨率低，因为只使用了四个红外传感器。对于第一个问题，我们构建了一个红外传感器的概率模型，该模型可以在不使用计算距离的情况下根据电压信息估计位置。第二个问题是通过始终水平旋转机器人来从各个方向获取信息来解决的。最后，利用电动转台和小车对定位性能进行了实验验证。在实验中，我们证实了即使机器人在运动中，即使机器人在靠近墙壁的地方飞行，定位也是成功的。少

In recent years, multirotor helicopter type autonomous UAVs are being used for aerial photography andaerial survey. In addition, various applications such as buildings maintenance, security and rescue are expected in multirotor helicopters. Not so distant future, these technology will be penetrate in our life. However, serious accident could occur if became widespread without guidelines and guarantees of safeness. Therefore, manufacturers need to be able to show that region of applicable application and behavior of fault mode to users. To show that, analytical model that can be used for analysis of behavioron model parameter change and hardware fault is needed. Our research group has started research onmultirotor helicopters since 2007. In 2010, we started to develop a practical multirotor with own design. Our research goal is to achieve advanced network system of ubiquitous small-scale aerial and groundrobots for high performance prevention of disaster in this project. In 2011, we suc … More ceeded in autonomouscontrol and we launched a consortium of multirotors in 2012. As described above, analytical model is needed for realizing safe multirotors. In this research, we constructed an analytical model which can beapply to general multirotors. The model was verified in experiment by constructing a control system. We constructed a simulation system which can be use for operator training and control system validation. Inaddition, this simulator can be simulated rotor failure and sensor failure. Actually by using this simulator, we showed that our simulator is useful for analyzing behavior of rotor failure, and is useful for training anoperator. While the multirotor helicopters are mainly used in outdoors, also it is expected for informationgathering in indoors. Autonomous navigation of flying robots in GPS-denied environments such as indoors requires that the the flying robot be able to estimate the position using external sensors. While laser scanners are mainly used for studies of indoor flight, development of smaller size robots is prevented due to the larger mass of sensor. Thus in this study, we develop a lightweight flying robot forachieving indoor autonomous flight using four infrared (IR) sensors. However, following problems are exists in localization based on IR sensors. First, it is difficult to use IR sensors close to a wall, because doing so would yield faulty results when calculating distance using the sensor output voltage. The secondproblem is that the spatial resolution is low because only four IR sensors are used. For the first problem, we constructed a probabilistic model of IR sensor that can be estimate position from voltage informationwithout the use of calculated distance. The second problem, was solved by rotating the robot horizontallyat all times to acquire information from various directions. Finally, the localization performance wasverified experimentally using an electric turntable and a cart. In the experiments, we confirmed that localization is successful even when the robot is in motion and even when the robot is flying near a wall. Less